The median time from whole blood collection and pancreatic cancer diagnosis was 6.3 years. MtDNA copy number was significantly different between cases and controls (p=0.04), with cases having a higher median mtDNA copy number (130.4; range 100-161.8 vs 122.2; range 100-147.1) (). Cases and controls did not significantly differ with respect to most characteristics; cases tended to have a greater history of diabetes, smoked for a longer time period, and consume more saturated fat and less carbohydrate (). The mean mtDNA copy number was 137.2 copies per cell for cases (standard deviation, 50; range, 60.5-362.9) and 126.9 copies per cell for controls (standard deviation, 37.3; range, 44.1-362.9).
shows the means and percentages of selected baseline characteristics among the controls according to quintile of mtDNA copy number. No clear, consistent trend existed for the selected baseline characteristics and mtDNA copy number. Men in the lowest quintile of mtDNA copy number tended to have smoked the most cigarettes for a greater number of years, and had the highest number of pack-years (P-trend=0.05).
MtDNA copy number was significantly associated with pancreatic cancer risk (highest mtDNA copy number quintile compared to lowest quintile, OR=1.64, 95%CI=1.01-2.67, p-trend=0.04) and in continuous models(OR=1.14, 95%CI=1.06-1.23) (). We also performed a dichotomous analysis by splitting mtDNA copy number at the median (OR=1.25, 95%CI=1.01-1.79). Similar associations were observed when comparing quartiles of mtDNA copy number(adjusted OR for 4th quartile versus 1st quartile=1.21, 95%CI=1.01-1.34; test for trend p-value=0.09) and when we restricted our analysis to the matched case-control sets and used conditional logistic regression to estimate risk (highest compared to lowest quintile, OR=1.23 , 95% CI=0.62-2.43 , continuous OR= 1.13, 95% CI=1.02-1.26 ). Patterns were similar when restricting the analysis to pancreas and lung controls only (Adjusted ORs for the 1st-5th quantiles in order=1.00, 0.76, 0.54, 0.94, 1.15; continuous OR=1.10,95%CI=1.01-1.19) and to pancreas and lymphoma controls only(Adjusted ORs for the 1st-5th quantile in order=1.00, 0.82, 0.63, 1.08,1.43; continuous OR=1.12,95%CI=1.04-1.22). The association between mtDNA copy number and pancreatic cancer was strongest in cases diagnosed during the first 7 years of follow-up (highest mtDNA copy number quintile OR= 2.14, 95% CI=1.16-3.96, p-trend=0.01, continuous OR=1.21, 95% CI 1.10-1.34, p=0.02), but not in cases occurring during follow-up greater than 7 years (highest mtDNA copy number quintile OR= 1.14, 95% CI=0.53-2.45, continuous OR=1.05, 95% CI 0.93-1.18).
Crude and Adjusted Odds Ratios and 95% Confidence Intervals for mtDNA Copy Number and Pancreatic Cancer Risk
Study intervention group did not significantly modify the association between mtDNA copy number and pancreatic cancer risk, although in adjusted models, positive associations appeared strongest among those randomized to the alpha-tocopherol (AT) supplementation group (continuous OR=1.30, 95%CI=1.09-1.57, p-trend=0.004; data not shown). However, there was no significant interaction by baseline serum alpha-tocopherol and serum alpha-tocopherol measured after the study intervention. The association between mtDNA copy number and pancreatic cancer was not modified by age, smoking habits, body mass index, physical activity, saturated fat, total energy, folate or carbohydrate intake, other dietary factors (antioxidants, fruit/vegetable intake), history of diabetes or gallstones, and physical activity levels (all P-values >0.05).